Private beta Join the waitlist
How it works

Each participant is built, briefed and kept apart.

A session rests on two things a single chat cannot give you: participants who each carry their own context, and a running order that decides when they are allowed to hear each other.

Four layers of context, per participant

  1. 01 PersonaTheir own Who they are: role, history, temperament, what they care about and what they are wary of. Their biases live here and nowhere else. Derek, for exampleSecurity lead who owns the SSO rollout. Went through a vendor breach in 2023 and has not forgotten it.
  2. 02 Shared experienceSet by you What the room has in common, when it should have anything in common: an employer, an event, a market. Nothing is shared unless you put it here. Derek, for exampleWorks at Northwind with the other three. Sat through last year’s rollout.
  3. 03 Knowledge baseTheir own The documents and facts this participant would actually have. Give the security lead the SOC 2 report and the CFO the invoice, and each argues from what they know. Derek, for exampleYour security questionnaire and SOC 2 report. Not the invoice, not the competitor’s quote.
  4. 04 Session memoryWhat was said aloud What they have heard in this session, plus their own private notes on it. Nothing from other sessions, and nothing another participant wrote privately. Derek, for exampleYour answer about the audit-log tier, and his own note: “Still no pen test.”

Isolation is enforced, and you can read exactly what was enforced

For every participant, the room panel lists what is in their context and what is kept out of it. Isolation is something you check rather than something we assert.

Derek Hollis context, round 2
In his context
  • His persona
  • Shared experience: Northwind, last year’s rollout
  • SOC 2 report, security questionnaire
  • Everything said aloud so far
  • His own private notes
Kept out
  • Priya’s and Marcus’s private notes
  • The invoice and the competitor’s quote
  • Your goal for the call
  • Any other session

The running order

Private rounds come before group rounds, in a fixed order. The order is not cosmetic: it is what makes the private answers usable as evidence.

When you take part

  1. 01 BriefSetup5 min You set the scenario and the cast, what the room shares and what each participant knows. You also write down what you want from it. No participant ever sees that. Participants who know what you want tend to give it to you.
  2. 02 Live roundYou take part10–20 min You speak as yourself. Participants answer from their own context, interrupt when their persona would, and ask what they would actually ask. Questions come from the participant who would ask them, in the order they would come.
  3. 03 Private reactionsIndividualeach turn After each of your turns, every participant writes a private note on how it landed. Captured as it happens and never shown to the rest of the room. The same answer lands four different ways. This is where you see it.
  4. 04 DebriefIndividual3 min Each participant gives you feedback in their own words: what worked, what did not, and what would have changed their mind. Every point links to the turn it is about. Four views, not an average of them.
  5. 05 Re-runOptional Run it again with the same cast and a different approach, then compare the private reactions turn by turn. The cast carries nothing over from the first run.

When you watch

  1. 01 Warm-upGroup2 min A short unrelated exchange so the room is talking before it is asked anything that matters. Discarded from the report. Cold first answers are worse answers.
  2. 02 First reactionIndividual3 min The same four questions to everyone, privately and at the same time: score out of ten, first thought, what they liked, what they did not. This is the round that makes the session evidence rather than a conversation.
  3. 03 Group discussionGroup12 min Moderated, with follow-ups that pick up what a participant actually said rather than working through a script. Scores move here. The report shows which participant moved them.
  4. 04 Rapid-fireExercise4 min Four short questions, answered privately and simultaneously after the discussion. Directly comparable to round 02. When this contradicts the discussion, the contradiction is the finding.
  5. 05 Wrap-upGroup3 min Each participant states their own condition for changing their mind, in their own words. The part product teams act on.

What this is not

The limits are part of the method. A tool that hid them would not be worth using.

Not a sample

Five participants are five participants. There is no population behind them, so there is no percentage to quote and the product never offers you one.

Not a substitute for the real conversation

It is what you do before you have earned a customer’s or a panel’s time, and how you decide what to ask or say once you have.

Not a prediction

A rehearsed sales call shows which objections are likely and how your answers land with each role. It does not tell you what the real buyer will decide.

Not hidden

Every participant is marked as AI-generated, every model is named, and each participant’s context is open for you to read.